An adaptive terahertz phased array beam tracking communication method and system

Abstract

The application discloses a kind of adaptive terahertz phased array beam tracking communication method and system, it is related to terahertz beam tracking technical field, lack of the prediction and fast dynamic response capability of high angular velocity target under the technical problem of mass codebook training, lack of adaptive perception scheduling capability of scanning resource of traversal scanning method.The application establishes the terahertz tracking-communication parallel transmission architecture of frequency division multiplexing decoupling, and transmitting end and receiving end both perform adaptive cross scanning, complete the establishment of initial beam;Adopt extended Kalman filter model to obtain the beam center of next time in combination with conical scanning, realize the adaptive guidance enhancement of scanning area;Scanning resource scheduling is carried out to the angular velocity state, in combination with the beam center of next time and scanning resource scheduling realizes beam tracking, while, cache history state is used for interruption discrimination and relocking recovery.The application has strong tracking capability in high dynamic scene, and the continuity and robustness of beam alignment are high.

Description

Technical Field

[0001] This invention belongs to the field of terahertz beam tracking technology, specifically relating to an adaptive terahertz phased array beam tracking communication method and system. Background Technology

[0002] With the increasing demands for Gbps-level throughput and low-latency communication from services such as low-altitude platform networks, vehicle-to-everything (V2X) networks, and high-speed data backhaul, communication frequency bands are continuously expanding towards millimeter-wave and terahertz bands. The terahertz band possesses abundant spectrum resources, and combined with phased array electronic scanning and high-gain beamforming capabilities, it has become an important technological path for achieving high-bandwidth mobile communication.

[0003] However, terahertz communication exhibits more pronounced narrow beamwidth, high path loss, and high directivity compared to traditional millimeter-wave communication. Due to the significantly increased path loss, both the transmitter and receiver often require large-scale, high-gain narrow beam arrays for dual-end alignment. This makes the commonly used "wide-beam-narrow-beam" cooperative tracking strategy in traditional millimeter-wave scenarios difficult to apply directly. The dual-end narrow beam configuration significantly expands the spatial scanning codebook size, causing traditional methods such as global exhaustive scanning, local neighborhood search, and fixed-step scanning to suffer from problems like excessively long scanning cycles, insufficient target dwell time, and link lock-loss issues in high-dynamic scenarios.

[0004] Existing beam tracking technologies attempt to incorporate Kalman filtering, position-assisted tracking, deep learning, or reinforcement learning methods. However, many of these methods are primarily geared towards millimeter-wave wide-beam scenarios, remaining at the simulation level or only optimizing a single tracking stage. They struggle to simultaneously address the initial acquisition efficiency, steady-state tracking accuracy, tracking resource overhead, and robustness of lock-down recovery for terahertz phased array systems under conditions of narrow beamwidths at both ends, high dynamics, and real-time service support. Compared to terahertz beam tracking schemes, current solutions typically lack the following capabilities: First, they lack system-level decoupling between the tracking beacon and the service transmission link, leading to increased system time-domain overhead; second, they lack the ability to predict and rapidly respond to high-angular-velocity targets under massive codebook training; third, traversal scanning methods lack adaptive sensing and scheduling capabilities for scanning resources; and fourth, they lack a complete closed loop for differentiating different interruption types and performing differentiated relock recovery based on historical state caching.

[0005] Furthermore, traditional beam training schemes that rely on time-division training or intra-frame pilot overhead consume valuable communication time-domain resources; when beam scanning occurs frequently, it further compresses the net throughput. In terahertz narrow-beam dynamic communication scenarios, if the tracking feedback is heavily dependent on the service signal itself, the feedback quality will deteriorate further during beam shift, deep fading, or momentary interruption of the service link, thereby reducing the stability of closed-loop control.

[0006] For example, the Chinese patent "Terahertz Communication Beam Tracking Method for High-Altitude Line-of-Sight Environments" (Publication No.: CN118316539A). This patent constructs a user movement model, generates the tracking range center and target physical direction angle set for the next frame based on the speed at multiple moments, and then obtains the delay parameters and subcarrier parameters by maximizing the power of the product of the received signal and the pilot, thereby determining the user's physical direction for the next frame.

[0007] However, this patented technology focuses more on achieving directional tracking of the user's mobility at one end, mainly targeting beam tracking scenarios where one end is fixed and the other end is moving. It does not address the alignment problem caused by the narrowing of the high-frequency beamwidth in the terahertz band, nor does it solve the problems of insufficient fast and accurate beam alignment and tracking and link interruption recovery capability when both the transmitting and receiving nodes in the terahertz band are moving, under narrow beam conditions. Summary of the Invention

[0008] This invention discloses a terahertz phased array adaptive beam tracking method and system, which aims to solve the technical problems of lack of prediction and fast dynamic response capability for high angular velocity targets under massive codebook training, and lack of adaptive sensing and scheduling capability for scanning resources in traversal scanning methods.

[0009] To solve the aforementioned technical problems, the present invention adopts the following technical solution:

[0010] An adaptive terahertz phased array beam tracking method includes the following steps:

[0011] Step S1: The terahertz phased array transmitter and receiver perform adaptive cross-scanning to complete the establishment of the initial beam at both ends;

[0012] Step S2: Establish an extended Kalman filter model based on the angular information of the relative directions between the receiver and transmitter;

[0013] Step S3: Combine the extended Kalman filter model and conical scanning to obtain the updated beam center;

[0014] Step S4: Extract the angular velocity state from the updated beam center in step S3, and realize the scanning resource scheduling of conical scanning based on the motion perception of the angular velocity state;

[0015] Step S5: Combine the beam center update in step S3 with the scan resource scheduling in step S4 to achieve beam tracking at the next moment.

[0016] Furthermore, the adaptive cross-scan in step S1 includes an initial capture phase and an iterative convergence phase;

[0017] Furthermore, in the initial acquisition phase, an initial scanning range, an initial step size, and a maximum number of iterations are preset. The transmitting end and the receiving end alternately perform blind scanning of the cross region in a preset order, and record the received signal strength indication value or beacon power value corresponding to each sampling beam.

[0018] Furthermore, in the iterative convergence phase, the optimal beam direction for the current round is determined based on the received signal strength indication value or the beacon power value. The optimal beam direction is judged according to the threshold to determine whether it meets the convergence condition. If it does not meet the condition, the scanning range and step size of the next round are reduced with the optimal beam direction as the center. The initial acquisition phase is repeated until the optimal beam direction meets the convergence condition, and the establishment of the dual-end initial beam is completed.

[0019] Furthermore, in step S2, the angular information of the relative direction between the receiver and the transmitter includes azimuth angle, azimuth angular velocity, pitch angle, and pitch angular velocity; the angular information serves as the state variable framework for the extended Kalman filter state vector.

[0020] Furthermore, the specific steps for obtaining the updated beam center in step S3 are as follows:

[0021] S31: Based on the state variables updated in the previous moment, the azimuth angle, elevation angle and their corresponding angular velocity state in the next moment are predicted by the extended Kalman filter model to obtain the candidate beam center; the state variables are the angle information updated in the previous moment.

[0022] S32: Using the candidate beam center obtained in step S31 as a reference, perform a conical scan around it to obtain the angle position of the maximum received power as the optimal power angle observation value.

[0023] S33: Input the candidate beam center in S31 and the optimal power angle observation in S32 into the update process of the extended Kalman filter model to obtain the updated beam center at the next time step.

[0024] Furthermore, in step S4, the angular velocity state is calculated based on the azimuth and elevation angles of the updated beam center; the scanning resource scheduling establishes a mapping relationship between the previous moment and the next moment of conical scanning through the angular velocity state, thereby realizing adaptive adjustment of the conical scanning parameters at the next moment.

[0025] Furthermore, in step S2, the extended Kalman filter model adopts a noise modeling strategy that combines fixed measurement noise with adaptive process noise, wherein the process noise covariance increases as the current estimated angular velocity increases;

[0026] Furthermore, the scanning resource scheduling in step S4 also includes adjusting the angular velocity according to multiple thresholds based on the conical scanning radius and the number of sampling points to form a multi-level scanning resource scheduling strategy under motion conditions.

[0027] An adaptive terahertz phased array beam tracking communication method includes the following steps:

[0028] The architecture is a frequency division multiplexing decoupled tracking-communication parallel terahertz phased array transmission architecture, in which the transmitter generates broadband communication signals for service transmission and narrowband beacon signals for beam tracking.

[0029] The receiving end performs power detection on the narrowband communication signal and uses the adaptive terahertz phased array beam tracking method for beam tracking.

[0030] Beam tracking status information is written to the historical status cache, and beam tracking interruption detection and relock recovery are performed based on the historical status cache;

[0031] The receiving end continuously performs baseband demodulation and service recovery on the broadband communication signal, realizing parallel operation of beam tracking and service transmission.

[0032] Furthermore, the beam tracking status information includes the updated beam center, angular velocity status, received signal strength, power fluctuation information, and corresponding time index for the current period; the interruption discrimination and relock recovery include the discrimination and recovery of obstruction-type interruptions and high-speed maneuver-type loss of lock.

[0033] An adaptive terahertz phased array beam tracking communication system, including

[0034] The phased array transceiver module transmits and receives terahertz signals and performs adaptive cross-scanning of the beam direction; the signal generation and frequency division multiplexing module generates narrowband beacon signals for tracking and broadband communication signals for service transmission; the power detection module detects the power of the narrowband beacon signals at the receiving end; the baseband processing module demodulates and restores the broadband communication signals; the beam control processing module performs adaptive cross-scanning, extended Kalman filter state estimation, beam center update, scan resource scheduling, and relock recovery; and the historical state buffer module caches historical beam state and power information. The beam control processing module controls the phased array transceiver module to update the beam direction based on the beacon power information output by the power detection module and operates in parallel with the baseband processing module. These modules cooperate to realize the adaptive terahertz phased array beam tracking communication method.

[0035] In summary, due to the adoption of the above technical solution, the beneficial effects of the present invention are:

[0036] 1. This invention achieves time-domain decoupling between the tracking closed loop and service transmission by transmitting narrowband tracking beacons and broadband service communication signals in parallel at different frequencies and receiving and processing them independently. This avoids the occupation of communication time slots by traditional time-division training methods, thereby reducing the impact of tracking on net throughput. Simultaneously, the independent beacon link avoids the risk of deep power fading of communication signals during beam scanning, ensuring the continuity and robustness of beam alignment.

[0037] 2. This invention does not only use extended Kalman filtering for beam direction estimation, but also uses the joint update of extended Kalman filter predictions and conical scan observations to achieve adaptive region guidance enhancement of the beam center, thereby improving the prediction and tracking capability of dual-end narrow beams in high dynamic scenarios.

[0038] 3. This invention further utilizes the angular velocity state output by the extended Kalman filter to achieve motion perception, and performs scanning geometry resource management accordingly. It adaptively and dynamically adjusts the conical scanning radius and the number of sampling points to form a scanning resource scheduling mechanism for different motion intensities, which can take into account both the anti-lockout capability under high-speed maneuvering conditions and the high gain performance during steady-state tracking.

[0039] 4. The present invention sets up an interruption discrimination and recovery mechanism based on historical state caching, which can distinguish between occlusion-type interruptions and high-speed maneuver-type lock loss, and adopt differentiated recovery strategies, which significantly improves the robustness of the link in complex dynamic scenarios.

[0040] 5. This invention forms a complete closed-loop beam management mechanism from initial acquisition and steady-state tracking to unlock recovery, which is suitable for terahertz phased array mobile tracking systems with narrow beams at both ends and has good engineering deployment value. Attached Figure Description

[0041] The present invention will be described by way of example and with reference to the accompanying drawings, wherein:

[0042] Figure 1 This is a flowchart of the terahertz phased array adaptive beam tracking method in Example 1.

[0043] Figure 2 This is the overall flowchart of the terahertz phased array adaptive beam tracking communication method in Example 2;

[0044] Figure 3 This is a diagram of the terahertz phased array tracking-communication system architecture based on frequency division multiplexing decoupling architecture in Example 2;

[0045] Figure 4 This is a diagram illustrating the indoor adaptive beam acquisition iterative process in Example 2;

[0046] Figure 5 This is a schematic diagram illustrating the optimal angle position iteration during the indoor adaptive beam tracking process in Example 2;

[0047] Figure 6 The above are measured results of beam tracking and communication under different outdoor motion conditions in Example 2. Detailed Implementation

[0048] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions in the embodiments of this application will be clearly and completely described below in conjunction with the embodiments and accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of the embodiments. The components of the embodiments of this application described and marked in the accompanying drawings can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of this application provided in the accompanying drawings is not intended to limit the scope of the claimed application, but merely represents selected embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without inventive effort are within the scope of protection of this application.

[0049] The following is combined with Figures 1-6 The present invention will be described in detail below.

[0050] Example 1

[0051] This embodiment discloses an adaptive terahertz phased array beam tracking method, referencing... Figure 1 , Figure 1 This is a flowchart of the terahertz phased array adaptive beam tracking method in this embodiment. The following is a summary of the process. Figure 1 Please provide an explanation.

[0052] The terahertz phased array adaptive beam tracking method specifically includes the following steps:

[0053] Step S1: The terahertz phased array transmitter and receiver perform adaptive cross-scanning to complete the establishment of the initial beam at both ends;

[0054] The adaptive cross-scan includes an initial capture phase and an iterative convergence phase;

[0055] The initial acquisition phase specifically involves: setting the initial scanning range, initial step size, and maximum number of iterations as conditions, and having the transmit and receive perform cross-scanning in the azimuth and elevation directions alternately in a preset order; in each round of cross-scanning, fixing the direction of one beam, and sequentially switching multiple scanning beams around the current candidate region at the other end, and recording the received signal strength indication value or beacon power value corresponding to each sampling beam.

[0056] The iterative convergence phase specifically involves: after completing one round of cross-scanning, extracting the beam direction corresponding to the maximum received signal strength as the optimal beam center estimate for the current round; determining whether the current step size is greater than the minimum step size threshold, whether the power improvement of the current round compared to the previous round is higher than the preset power improvement threshold, and whether the current maximum power is higher than the minimum power threshold; if the conditions for continuing the search are met, then the range and step size of the next round of cross-scanning are reduced with the current optimal beam center as the center, and the initial acquisition phase is repeated; if the convergence conditions are met, then the dual-end initial acquisition and beam handshake are completed.

[0057] Step S2: Establish an extended Kalman filter model based on the angular information of the relative directions between the receiver and transmitter;

[0058] The angular information between the receiver and transmitter relative to each other includes: the azimuth angle, azimuth angular velocity, elevation angle, and elevation angular velocity after acquisition. This angular information is used as the extended Kalman filter state vector.

[0059] The extended Kalman filter model parameters are initialized based on the extended Kalman filter state vector. Specifically, the optimal beam center obtained during the initial acquisition phase is used as the initial angle state, and the angular velocity value estimated from adjacent moments or a preset value is used as the initial angular velocity state. The state covariance matrix and filter parameters are initialized using a diagonal matrix. The optimal beam center consists of azimuth and elevation angles. The filter parameters include state transition model parameters, sampling period, observation model parameters, observation matrix H, and noise parameters.

[0060] A state evolution equation based on a constant velocity motion model is used to predict the state vector at the current moment, thus obtaining the predicted beam center. Multiple sampling points are obtained through adaptive conical scanning, and the overall power field distribution is acquired from these sampling points. This result is then weighted to obtain a new beam center, whose corresponding angle is used as the observation vector at the current moment. An observation model based on linear mapping is used to establish the relationship between the observation vector and the state vector. According to the rules for updating the residual between the observation vector and the predicted state vector, the prior predicted state vector is updated to obtain the posterior state vector at the current moment.

[0061] In another embodiment, the extended Kalman filter employs a noise modeling strategy combining fixed measurement noise and adaptive process noise. The measurement noise characterizes the measurement accuracy of the conical scan observation, while the process noise is dynamically adjusted based on the current angular velocity state to characterize the deviation of the target's true motion from the constant velocity model. This maintains good smoothness in static scenes and improves predictive response capability in highly dynamic scenes. The adaptive process noise can be:

[0062]

[0063] in, As the reference noise for static scenes, For the current estimated angular velocity amplitude, A matrix is ​​selected for the velocity state. Using the above method, when the target transitions from stationary to rapid motion, the Q value automatically increases, thereby improving the filter's response speed to dynamic changes.

[0064] Step S3: Combine the extended Kalman filter model and conical scanning to obtain the updated beam center;

[0065] Step S3 is used to obtain the updated beam center for the next moment based on the state variables of the previous moment, combined with the extended Kalman filter model and conical scanning, specifically as follows:

[0066] S31: Based on the state variables updated in the previous moment, the azimuth angle, elevation angle and their corresponding angular velocity state in the next moment are predicted by the extended Kalman filter model to obtain the candidate beam center; the state variables are the angle information updated in the previous moment.

[0067] At the start of the current tracking cycle, based on the updated state variables and state transition model from the previous moment, the azimuth angle, elevation angle, and their corresponding angular velocity state for the next moment are predicted to obtain the candidate beam center. This candidate beam center is not directly used as the final tracking result, but rather as the reference center for subsequent scanning area settings to achieve adaptive area guidance enhancement of the scanning area.

[0068] The state prediction stage and its prediction covariance can be expressed as:

[0069]

[0070]

[0071] in, Represents the state at different times. Represents the state evolution equation. Represents the predicted covariance. The representative process noise covariance.

[0072] S32: Using the candidate beam center obtained in step S31 as a reference, perform a conical scan around it to obtain the angle position of the maximum received power as the optimal power angle observation value.

[0073] Using the candidate beam center as a reference, a conical scan is performed around it to obtain the received signal intensity at multiple sampling points; based on the comparison results of the power at multiple sampling points, the corresponding received power field distribution within the current tracking period is extracted, and the centroid position is obtained by weighting as the optimal beam center, and this angular position is used as the optimal power angle observation value;

[0074] During the conical scanning process, in the k-th tracking period, let the candidate beam center predicted by the extended Kalman filter be... :

[0075]

[0076] in, To predict the azimuth, To predict the pitch angle.

[0077] Let the current conical scan radius be... The number of sampling points is The scanning phase angle of the i-th sampling point is defined as:

[0078]

[0079] The direction of the i-th sampling beam is:

[0080]

[0081]

[0082] The received power measured at this sampling point is:

[0083]

[0084] First, the power at each sampling point is normalized. Let... , This is the minimum power among all sampling points in the current period, used to remove background bias.

[0085] Then define the normalized weights:

[0086]

[0087] in, To prevent positive numbers with a denominator of zero.

[0088] The observation offset obtained from the conical scan power is:

[0089]

[0090]

[0091] The new observation center utilizes the local power field distribution information of the entire conical sampling surface to construct a new weighted centroid beam center.

[0092] The corresponding observation vector is represented as

[0093]

[0094] S33: Input the candidate beam center in S31 and the optimal power angle observation value in S32 into the update process of the extended Kalman filter model to obtain the updated beam center;

[0095] The candidate beam center and the optimal power angle observation are input into the update process of the extended Kalman filter model to obtain the updated state at the current moment; wherein, the updated azimuth and elevation angles together constitute the updated beam center at the current moment. In this step, the extended Kalman filter model is not only used for angle state estimation, but also achieves adaptive region guidance enhancement for subsequent scanning areas by combining the predicted value with the conical scan observation value for the current beam center update.

[0096] The stage of updating the beam center by fusing predicted and observed values ​​is specifically as follows:

[0097] The residual information from the predicted and actual observations is expressed as:

[0098]

[0099] in, Represents the observation vector. Representative observation equation

[0100] To describe the observational uncertainties, including prediction of propagation and measurement noise, the information covariance is:

[0101]

[0102] This represents the observation noise covariance.

[0103] The Kalman gain is:

[0104]

[0105] The beam center update step combines historical beam center data with the optimal beam center during actual tracking to estimate the next center in advance, avoiding blind conical scanning and improving efficiency. The updated and corrected beam state and covariance based on prediction and actual observation results are expressed as follows:

[0106]

[0107]

[0108] in, Represents the predicted beam state. Represents the identity matrix.

[0109] After updating the extended Kalman filter based on the predicted state and the observed vector, the posterior state vector for the k-th tracking period is obtained:

[0110]

[0111] And obtain the current corrected beam center based on the updated posterior state:

[0112]

[0113] Based on this, the current corrected beam center and angular velocity state will continue to be used for the next cycle of scanning area guidance, and the guidance center for the next cycle will be defined as:

[0114]

[0115] in, This is the velocity advance coefficient. This is the residual compensation coefficient. Sampling time.

[0116] Step S4: Extract the angular velocity state from the updated beam center in step S3, and realize the scanning resource scheduling of conical scanning based on the motion perception of the angular velocity state;

[0117] The angular velocity state is extracted from the beam center updated in step S3 and used to characterize the target motion intensity. Scanning resource scheduling is performed based on the angular velocity state, which is calculated from the azimuth and elevation angles.

[0118]

[0119] In this embodiment, the conical scanning radius and the number of sampling points are dynamically adjusted according to the continuous mapping relationship, and the scanning resource scheduling is as follows:

[0120]

[0121]

[0122] in, Indicates the current conical scan radius. Indicates the current number of sampling points. This represents the radius of the conical scan at the previous moment. This indicates the number of sampling points at the previous time step. and The adjustment coefficient is determined based on the angular velocity state and is obtained from historical conical scanning experience.

[0123] In another embodiment, the conical scanning radius and the number of sampling points can be adjusted in stages according to multiple angular velocity thresholds. When the angular velocity is higher than a preset high-speed threshold, the conical scanning radius and the number of sampling points in the next tracking cycle are increased to expand the scanning coverage and prevent loss of lock. When the angular velocity is lower than a preset low-speed threshold, the conical scanning radius and the number of sampling points are decreased to reduce scanning disturbances and improve steady-state tracking gain. When the angular velocity is in the middle range, the corresponding middle radius and number of sampling points combination is used. In this step, the angular velocity is not only used for motion state characterization, but also for adaptive adjustment of the conical scanning radius and the number of sampling points, thereby realizing scanning resource scheduling based on extended Kalman filter motion sensing.

[0124] Step S5: Combine the beam center update in step S3 with the scanning resource scheduling in step S4 to achieve beam tracking at the next moment;

[0125] Based on the updated beam center obtained in step S3 and the conical scanning parameters obtained in step S4, beam control commands for the local and / or remote phased arrays are generated, and the system switches to the new transmit and receive beams.

[0126] Example 2

[0127] This embodiment discloses an adaptive terahertz phased array beam tracking communication method, referencing... Figure 2 , Figure 2 This is a flowchart of the adaptive terahertz phased array beam tracking communication method in this embodiment, which includes the following steps:

[0128] The architecture is a frequency division multiplexing decoupled tracking-communication parallel terahertz phased array transmission architecture, in which the transmitter generates broadband communication signals for service transmission and narrowband beacon signals for beam tracking.

[0129] The receiving end performs power detection on the narrowband communication signal and uses the above-mentioned adaptive terahertz phased array beam tracking method for beam tracking.

[0130] Beam tracking status information is written to the historical status cache, and beam tracking interruption detection and relock recovery are performed based on the historical status cache;

[0131] The receiving end continuously performs baseband demodulation and service recovery on the broadband communication signal, realizing parallel operation of beam tracking and service transmission.

[0132] The broadband communication signal and the narrowband beacon signal have different center frequencies and are transmitted by a terahertz phased array after frequency variation. The receiving end performs high-speed power detection on the narrowband beacon signal and performs independent demodulation processing on the broadband communication signal to make the beam tracking closed loop and service transmission independent of each other in the time domain.

[0133] The historical state recording cache specifically involves writing the updated beam center, angular velocity state, received signal strength, power fluctuation information, and corresponding time index of the current period into the historical state cache at the end of each tracking cycle.

[0134] In another embodiment, the historical state cache can save state and power information from the most recent tracking cycles to balance recovery effectiveness and storage overhead.

[0135] The interruption includes obstruction-type interruption and high-speed maneuvering-type loss of lock. The interruption is determined based on the historical state buffer, specifically: the received signal strength is detected in real time, and the received power change information and angular velocity status in the historical state buffer are read; when the received power drops suddenly and the angular velocity status is small or does not change significantly, it is determined to be an obstruction-type interruption; when the received power drops and the angular velocity status is large or shows a rapid changing trend, it is determined to be a high-speed maneuvering-type loss of lock.

[0136] The relock recovery includes occlusion-type interruption recovery and high-speed maneuvering-type loss-of-lock recovery. The occlusion-type interruption recovery maintains target path prediction and continuously updates the recovery anchor point by calling the backward state sequence in the historical state cache. After detecting that the occlusion has been removed or the received power has increased again, a small-range fine scan is performed with the recovery anchor point as the center until the main lobe lock is restored. The high-speed maneuvering-type loss-of-lock recovery performs forward prediction by calling the angle state and angular velocity state in the historical state cache before the interruption to estimate the possible new position of the target, and performs a hierarchical scan with the predicted position as the anchor point. The hierarchical scan includes first performing a coarse-range scan to expand the acquisition area, and then performing a fine-range scan to restore the main lobe lock precisely.

[0137] In another embodiment, the recovery anchor point in the relock recovery can be determined by the historical best beam center, the forward predicted beam center, or a combination of both.

[0138] If, after recovery via relocking, the received power returns to the normal tracking power range and the beam center stabilizes again, the current beam center and historical state cache are updated, and steady-state tracking continues. If recovery fails, the search range is expanded, some scanning parameters are reset, and adaptive cross-scan initial acquisition is re-executed.

[0139] In this embodiment, the beam tracking performance of the adaptive terahertz phased array beam tracking communication method disclosed in this invention is verified through experiments. The following is a summary of the verification process. Figures 3 to 6 To provide further explanation.

[0140] The terahertz phased array adaptive beam tracking system employs 64-channel phased array modules with an 8×8 array size at both the transmitter and receiver. The array has a 3 dB beamwidth of approximately 10° and a scanning range of ±30°. The system uses a frequency division multiplexing decoupling architecture, with the communication signal center frequency at 93.2 GHz and a bandwidth of 0.8 GHz, while the tracking beacon is a 94 GHz narrowband continuous wave signal.

[0141] The transmitting end includes: a beam management CPU system, baseband modulation board, signal source, intermediate frequency mixer, phased array transmitter module, and service input interface based on the Wildfire LubanCat 5 high-performance embedded development board with Rockchip RK3588 SoC; the receiving end includes: a beam management CPU system, phased array receiver module, RF power detection module, baseband demodulation board, beam control processor, and display output interface.

[0142] The connection methods between the various modules are as follows Figure 3 As shown, Figure 3 This embodiment presents a terahertz phased array tracking-communication system architecture based on a frequency division multiplexing decoupling architecture. In this system, the baseband board generates a single-carrier communication service signal, and the signal source generates a narrowband continuous wave tracking beacon. Both signals are converted to intermediate frequency and then input into the phased array module for transmission. The receiving end performs power detection on the 94 GHz beacon and demodulates and recovers the 93.2 GHz communication signal.

[0143] When the system starts, both communicating parties first enter an adaptive cross-scanning phase. Cross-scanning is performed on the azimuth and elevation angles using a preset scanning range and initial step size. The transmitting and receiving ends alternate sequentially. Each scan records the beam direction corresponding to the beacon's maximum received power, and uses this as the center for the next scan to narrow the search area. After several iterations, a two-end beam handshake is completed, and an initial link is established. Figure 4 As shown, Figure 4 This represents the feedback RSSI power iteration, iteration step, and scan range iteration during the beam scanning acquisition phase. As the number of iterations increases, both the power and scan range converge, and the power eventually stabilizes. When the power reaches -39 dBm, the beam handshake is completed and the initial link is established.

[0144] In an indoor validation process, the total initial scan time can be controlled within the range of approximately milliseconds to seconds. Simulation statistics show that the average initial acquisition time is approximately 960 ms.

[0145] After initial acquisition, the system enters the adaptive conical tracking phase based on extended Kalman filter motion sensing. In this embodiment, the extended Kalman filter state variables are azimuth angle, azimuth angular velocity, elevation angle, and elevation angular velocity; the optimal power angle is output for each cycle of the conical scan and used as the observation input to the extended Kalman filter; the extended Kalman filter prediction value and the optimal power angle observation value are jointly used to update the beam center at the next moment, forming adaptive guidance for the subsequent scanning area. In the indoor dynamic experiment, the transmitter is mounted on a gimbal, and the target azimuth change is simulated by setting trajectory points; the maximum angular velocity can reach approximately 10° / s. Experimental results show that the system can maintain stable main lobe tracking within this dynamic range and maintain small power fluctuations.

[0146] During steady-state tracking, the beam control processor determines whether the target is in a low-dynamic, medium-dynamic, or high-dynamic state based on the angular velocity state output by the extended Kalman filter. When the target is in a low-dynamic or static state, a smaller cone radius and fewer sampling points are used to reduce the disturbance to the main lobe gain of the communication; when the target is in a high-dynamic state, the cone radius and the number of sampling points are increased to expand the search coverage and reduce the probability of loss of lock.

[0147] When the receiver detects that the beacon power is consistently below the threshold and the power fluctuation meets the interruption criteria, the system invokes the historical state cache module. Simulation statistics show that in obstruction-type interruption scenarios, the recovery success rate can reach approximately 99%; in high-speed maneuvering-type interruption scenarios, the relock success rate within 2 seconds can reach approximately 97%.

[0148] In an outdoor verification process, the receiver was deployed at a fixed location by a lake, while the transmitter was mounted on a vehicle moving approximately perpendicular to the line-of-sight distance, with a transmission and reception distance of about 900 m and a vehicle speed of about 20 km / h. The system uses a 1.9 GHz continuous wave intermediate frequency beacon and a 1.1 GHz center frequency communication intermediate frequency signal, which are up-converted to form a 94 GHz beacon and a 93.2 GHz communication signal, respectively. The service data is a 1080p 30 fps real-time video stream of about 1.5 Gbps. The beam control algorithm runs on an embedded processing platform, with a single-beam closed-loop refresh latency of about 5 ms.

[0149] refer to Figure 5 , Figure 5 The diagram illustrates the optimal angle change after six iterations in the beam tracking phase. Experimental results show that the system can achieve stable main lobe tracking within an approximately 15° azimuth range during vehicle movement. Figure 6 , Figure 6This represents the specific beam tracking process feedback RSSI received power change process during the beam tracking phase, demonstrating beam tracking performance testing under conditions of near-interruption and recovery due to rapid maneuvers during the switching between stationary, tracking, and stationary states; combined with Figure 6 It can be seen that when an interruption occurs and is restored, the received power fluctuation can be suppressed to a small range, and the communication link maintains a block error rate of less than 1% in dynamic scenarios. It can stably transmit real-time uncompressed video services. This result verifies the effectiveness of the method and system proposed in this invention in high-dynamic terahertz phased array communication scenarios.

[0150] Example 3

[0151] This embodiment discloses an adaptive terahertz phased array beam tracking communication system. The system is set on both communication parties, with each party serving as the local end and the peer end. It is used to execute the terahertz phased array adaptive beam tracking method and includes: a phased array transceiver module, a signal generation and frequency division multiplexing module, a power detection module, a baseband processing module, a beam control processing module, and a historical state cache module.

[0152] The phased array transceiver module is used to transmit and receive terahertz signals and perform electronically controlled scanning of the beam direction; the signal generation and frequency division multiplexing module is used to generate narrowband beacon signals for tracking and broadband communication signals for service transmission; the power detection module is used to detect the power of the narrowband beacon signals at the receiving end; the baseband processing module is used to demodulate and restore the broadband communication signals; the beam control processing module is used to perform adaptive cross-scanning, extended Kalman filter state estimation, beam center update, scan resource scheduling, and relock recovery; the historical state cache pool module is used to cache historical beam state and power information; the beam control processing module controls the phased array transceiver module to update the beam direction based on the beacon power information output by the power detection module, and runs in parallel with the baseband processing module;

[0153] In another embodiment, the beam control processing module includes: an initial acquisition unit, a state prediction unit, a conical scanning unit, a beam center update unit, a resource scheduling unit, an interruption discrimination unit, and a recovery control unit; the initial acquisition unit is used to perform cross-scanning and two-end beam handshake; the state prediction unit is used to establish and update the extended Kalman filter state vector; the conical scanning unit is used to perform adaptive conical scanning around the candidate beam center to obtain the received power field distribution, and weighted to obtain the centroid position as the optimal beam center, i.e., the optimal power angle observation value; the beam center update unit is used to fuse the predicted value and the observed value to update the current beam center; the resource scheduling unit is used to adjust the conical scanning radius and the number of sampling points according to the angular velocity state; the interruption discrimination unit is used to determine the link interruption type based on the historical state buffer and received power information; the recovery control unit is used to execute corresponding relocking recovery strategies for obstruction-type interruptions and high-speed maneuvering-type loss of lock, respectively.

[0154] The above description of the disclosed embodiments enables those skilled in the art to make or use the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. An adaptive terahertz phased array beam tracking method, characterized in that: Includes the following steps: Step S1: The terahertz phased array transmitter and receiver perform adaptive cross-scanning to complete the establishment of the initial beam at both ends; Step S2: Establish an extended Kalman filter model based on the angular information of the relative directions between the receiver and transmitter; Step S3: Combine the extended Kalman filter model and conical scanning to obtain the updated beam center; Step S4: Extract the angular velocity state from the updated beam center in step S3, and realize the scanning resource scheduling of conical scanning based on the motion perception of the angular velocity state; Step S5: Combine the beam center update in step S3 with the scan resource scheduling in step S4 to achieve beam tracking at the next moment.

2. The adaptive terahertz phased array beam tracking method according to claim 1, characterized in that: The adaptive cross-scan in step S1 includes an initial capture phase and an iterative convergence phase; The initial acquisition phase presets the initial scanning range, initial step size and maximum number of iterations. The transmitter and receiver alternately perform blind scanning of the cross region in a preset order, and record the received signal strength indication value or beacon power value corresponding to each sampling beam. The iterative convergence phase determines the optimal beam direction for the current round based on the received signal strength indication value or beacon power value. It then determines whether the optimal beam direction meets the convergence condition based on a threshold. If not, it narrows the scanning range and step size for the next round, using the optimal beam direction as the center. The initial acquisition phase is repeated until the optimal beam direction meets the convergence condition, thus completing the establishment of the dual-end initial beam.

3. The adaptive terahertz phased array beam tracking method according to claim 1, characterized in that: In step S2, the angle information between the receiver and the transmitter relative to each other includes azimuth angle, azimuth angular velocity, pitch angle, and pitch angular velocity; the angle information serves as the state variable framework for the extended Kalman filter state vector.

4. The adaptive terahertz phased array beam tracking method according to claim 1, characterized in that: The specific steps for obtaining the updated beam center in step S3 are as follows: S31: Based on the state variables updated in the previous moment, the azimuth angle, elevation angle and their corresponding angular velocity state in the next moment are predicted by the extended Kalman filter model to obtain the candidate beam center; the state variables are the angle information updated in the previous moment. S32: Using the candidate beam center obtained in step S31 as a reference, perform a conical scan around it to obtain the angle position of the maximum received power as the optimal power angle observation value. S33: Input the candidate beam center in S31 and the optimal power angle observation value in S32 into the update process of the extended Kalman filter model to obtain the updated beam center at the next time step.

5. The adaptive terahertz phased array beam tracking method according to claim 1, characterized in that: In step S4, the angular velocity state is calculated based on the azimuth and elevation angles of the updated beam center; the scanning resource scheduling establishes a mapping relationship between the previous moment and the next moment of conical scanning through the angular velocity state, thereby realizing adaptive adjustment of the conical scanning parameters at the next moment.

6. The adaptive terahertz phased array beam tracking method according to claim 1, characterized in that: In step S2, the extended Kalman filter model adopts a noise modeling strategy that combines fixed measurement noise with adaptive process noise. The process noise covariance increases as the current estimated angular velocity increases.

7. The adaptive terahertz phased array beam tracking method according to claim 1, characterized in that: The scanning resource scheduling in step S4 also includes adjusting the angular velocity according to multiple thresholds based on the conical scanning radius and the number of sampling points to form a multi-level scanning resource scheduling strategy under motion conditions.

8. An adaptive terahertz phased array beam tracking communication method, characterized in that: Includes the following steps: The architecture is a frequency division multiplexing decoupled tracking-communication parallel terahertz phased array transmission architecture, in which the transmitter generates broadband communication signals for service transmission and narrowband beacon signals for beam tracking. The receiving end performs power detection on the narrowband communication signal and performs beam tracking using the adaptive terahertz phased array beam tracking method described in any one of claims 1 to 7. Beam tracking status information is written to the historical status cache, and beam tracking interruption detection and relock recovery are performed based on the historical status cache; The receiving end continuously performs baseband demodulation and service recovery on the broadband communication signal, realizing parallel operation of beam tracking and service transmission.

9. The adaptive terahertz phased array beam tracking communication method according to claim 8, characterized in that: The beam tracking status information includes the updated beam center, angular velocity status, received signal strength, power fluctuation information, and corresponding time index for the current period; the interruption discrimination and relock recovery include the discrimination and recovery of obstruction-type interruption and high-speed maneuvering-type loss of lock.

10. An adaptive terahertz phased array beam tracking communication system, characterized in that: include The phased array transceiver module is used to transmit and receive terahertz signals and perform adaptive cross-scanning of the beam direction; the signal generation and frequency division multiplexing module is used to generate narrowband beacon signals for tracking and broadband communication signals for service transmission; the power detection module is used to perform power detection on the narrowband beacon signals at the receiving end. The baseband processing module is used for demodulating and restoring broadband communication signals; the beam control processing module is used for performing adaptive cross-scanning, extended Kalman filter state estimation, beam center update, scan resource scheduling, and relocking recovery; the historical state cache pool module is used for caching historical beam state and power information; the beam control processing module controls the phased array transceiver module to update the beam pointing based on the beacon power information output by the power detection module, and runs in parallel with the baseband processing module; the various modules cooperate with each other to realize the adaptive terahertz phased array beam tracking communication method according to any one of claims 8 or 9.

Images